1. CD40
CD40 is an antigen with a molecular weight of 50 kDa that is present on the cell membrane surface. CD40 is expressed on B cells, dendritic cells (DC), certain types of cancer cells, and thymic epithelial cells. CD40 is known to play a key role in proliferation and differentiation of B cells and DC. CD40 has been identified as an antigen that is expressed on the human B cell surface (E. A. Clark et. al., Proc. Natl. Acad. Sci. USA 83: 4494, 1986, I. Stamenkovic et. al., EMBO J. 8:1403, 1989). Based on the amino acid sequence homology, CD40 is thought to be a member of the TNF receptor family, to which a low affinity NGF receptor, TNF receptor, CD27, OX40, CD30 and the like belong. The gene of a ligand (CD40L) for human and mouse CD40 has been cloned recently, revealing that it is a type II membrane protein, and is expressed on activated CD4+T cells. It has also been shown that CD40L introduces strong activation signals into human and mouse B cells.
The expression of CD40 has been confirmed more often on dendritic cells than on B cells, so that it has become clear that CD40 plays an important role. The binding of CD40 with CD40L causes the activation of antigen-presenting cells (APC). Specifically, it enhances the expression of co-stimulation molecules such as CD80 (B7-1) and CD86 (B7-2), or the production of IL-12 (Caux, C., et al.: Activation of human dendritic cells through CD40 cross-linking. J. Exp. Med., 180:1263, 1994), (Shu, U., et al: Activated T cells induce interleukin-12 production by monocyte via CD40-CD40 ligand interaction. Eur. J. Immunol. 25: 1125, 1995). Dendritic cells show strong antigen-presenting ability, and have strong helper T (Th) cell-activating ability. Furthermore, it is thought that dendritic cells control the differentiation of naive Th cells into Th1 or Th2 cells. When dendritic cells (DC1) are made to mature by culturing peripheral blood monocytes that are myeloid dendritic cells in the presence of GM-CSF and IL-4 and using CD40L, the DC1 in vitro are capable of producing IL-12, stimulate and activate allo-naive Th cells, and thus induce IFNγ-producing T cells (specifically, promotes differentiation into Th1). Since this action is inhibited by anti-IL-12 antibodies, the reaction may be mediated by IL-12. On the other hand, when lymphocyte-dendritic cells (DC2) are prepared by culturing lymphatic tissue T regions or plasmacytoid T cells present in peripheral blood with IL-3 and CD40 ligands, DC2 are incapable of producing IL-12, stimulate and activate allo-naive Th cells, induce IL-4-producing T cells, and thus promote differentiation into Th2. It is thought that Th1 cells are involved in the activation of cellular immunity, and Th2 cells are involved in enhancement of the ability for humoral immunity as well as the suppression of the ability for cellular immunity. Cytotoxic T cells (CTL) activated with the help of Th1 cells can remove causative factors (many viruses, Listeria monocytogenes, tubercle bacillus, toxoplasma protozoa and the like) multiplying in the cytoplasm and tumor cells.
It has been shown that anti-CD40 monoclonal antibodies that recognize CD40 expressed on the membrane surfaces exert a variety of biological activities on B cells. Anti-CD40 monoclonal antibodies are largely classified into agonistic and antagonistic antibodies impacting the interaction between CD40 and CD40L.
2. Agonistic Antibody
The activation of B cells is known as an action of agonistic antibodies. For example, anti-CD40 antibodies have been reported to induce cell adhesion (Barrett et al., J. Immunol. 146: 1722, 1991; Gordon et al., J. Immunol. 140: 1425, 1988), enhance cell size (Gordon et al., J. Immunol. 140: 1425, 1988; Valle et al., Eur. J. Immunol. 19: 1463, 1989), induce the division of B cells that are activated only with anti-IgM antibodies, anti-CD20 antibodies or phorbol ester (Clark and Ledbetter, Proc. Natl. Acad. Sci. USA 83: 4494, 1986; Gordon et al., LEUCOCYTE TYPING III. A. J. McMicheal ed. Oxford University Press. Oxford, p. 426; Paulie et al., J. Immunol. 142: 590, 1989), induce the division of B cells in the presence of IL4 (Valle et al., Eur. J. Immunol. 19: 1463, 1989; Gordon et al., Eur. J. Immunol. 17: 1535, 1987), induce the expression of IgE (Jabara et al., J. Exp. Med. 172: 1861, 1990; Gascan et al., J Immunol. 147: 8, 1991), IgG and IgM (Gascan et al., J. Immunol. 147: 8, 1991) of cells stimulated with IL-4 and cultured without T cells, enhance the secretion and the on-the-cell expression (Challa A, Allergy, 54: 576, 1999) of soluble CD23/Fcε RII from B cells by IL-4 (Gordon and Guy, Immunol. Today 8: 339, 1987; Cairns et al., Eur. J. Immunol. 18: 349, 1988), and promote IL-6 production (Clark and Shu, J. Immunol. 145: 1400, 1990). Furthermore, it has been reported that B cell clones are established from human primary culture B cells by adding IL-4 and anti-CD40antibodies in the presence of CDw32+ adhesion cells (Bancherau et al., Science 241:70,1991), and the inhibition of the apoptosis of germinal center cells is mediated by CD40, regardless of the function of antigen receptors (Liu et al., Nature 342: 929,1989). As described above, CD40 has been identified as an antigen expressed on the human B cell surface. Thus, most of the isolated antibodies have been evaluated mainly using function to induce the proliferation and differentiation of human B cells and activity to induce cell death in cancer cells as indicators (Katira, A. et. al., LEUKOCYTE TYPING V. S. F. Schlossossman, et al. eds. p. 547. Oxford University Press. Oxford, W. C. Flansow et al., LEUKOCYTE TYPING V. S. F. Schlossossman, et al. eds. p. 555. Oxford University Press. Oxford, J. D. Pound et al., International Immunology, 11: 11, 1999).
Anti-CD40 antibodies were shown to cause the maturation of DC (Z. H. Zhou et. al., Hybridoma, 18: 471 1999). Moreover, the role of CD4T cells in antigen-specific CD8T cell priming has been reported to activate DC via CD40-CD40L signaling. It was shown that the role of CD4 helper T cells in activation of dendritic cells (DC) can be replaced by that of anti-CD40 monoclonal antibodies (mAb) (Shoenberger, S. P., et al.: T-cell help for cytotoxic T lymphocytes is mediated by CD40-CD40L interactions. Nature, 480, 1998). Furthermore, it was shown in mice that the organism can be protected not only from tumor cells expressing CD40 but also from tumor cells not expressing the same by the administration of anti-CD40 antibodies (French, R. R., et. al.: CD40 antibody evokes a cytotoxic T-cell response that eradicates lymplioma and bypasses T-cell help. Nature Medicine, 5, 1 999).
Most antibodies reported to date have not been isolated using the effect on DC as an indicator. However, in terms of the modification of DC functions, antibodies selected by their action on B cells are likely to be insufficient as therapeutic agents. It was reported that among monoclonal antibodies against mouse CD40, there are clones that react to DC, but do not react to vascular endothelial cells, and, conversely, clones that do not react to DC, but react to vascular endothelial cells, depending on epitopes that the antibodies recognize (Van Den Berg, TK, et. al., Immunology, 88: 294, 1996). It is also assumed that the binding and action of human CD40 antibodies to DC differ depending on epitopes.
It is known that anti-CD40 antibodies or CD40 ligands can suppress the proliferation of CD40-expressing lymphoma cell lines and thus can induce the cell death (Funakoshi S et al., Blood, 83: 2782, 1994; Funakoshi S et al., Journal of Immunotherapy, 19, 93, 1996; Z. H. Zhou et. al., Hybridoma, 18: 471 1999; and Joseph A et al., Cancer Research, 60: 3225, 2000). What is interesting about agonistic antibodies is that the function of the antibody does not always coincide always with that of CD40L. Action to activate B cells does not also coincide with action to suppress B cell tumor growth. It is desired to develop antibodies having both DC-activating ability and tumor cell proliferation-suppressing action. Moreover, among agonistic antibodies, both antibodies that inhibit and those that do not inhibit the binding of CD40L to CD40 are present (Challa A et al., Allergy, 54: 576, 1999). For example, antibodies produced by G28-5 (ATCC No.HB-9110) compete with CD40L, so that there is no effect resulting from the combined use with CD40L. The degree of activation of CD40-expressing cells differs depending on antibodies. Even when antibodies exhibit independently weak agonistic activity, the combined use of the antibodies with CD40 ligands may more significantly promote the activity in the presence of the antibodies, than the activity resulting from CD40 ligands alone. In contrast, even when antibodies exhibit independently agonistic activity, inhibition of CD40 ligands may lower the activity in the presence of the antibodies to a greater extent than the activity resulting from CD40 ligands alone (Pound et al., International Immunology, 11:11, 1999). It was shown that with antibodies that do not compete with CD40 ligands, stronger suppression of proliferation can be achieved in the presence of CD40 ligands, although the tumor cell proliferation-suppressing action of the antibody itself is weak (Joseph A et al., Cancer Research, 60: 3225, 2000). Accordingly, it is desired to develop antibodies that bind to CD40 to suppress independently cell proliferation, but that do not inhibit the binding of CD40 ligands to CD40. By taking full advantage of such characteristics, there is a possibility of developing a therapeutic agent that is more efficient than a soluble CD40L. For example, the soluble CD40L is activated by binding with CD40, and at the same time, it suppresses the function of CD40L present in vivo. An antibody that does not compete with CD40L, does not cause such suppression, and has better therapeutic effects can be expected by synergistic effect.
3. Antagonistic Antibody
In the meantime, as described above, it is expected that because CD40 plays an important role in immune reaction, therapeutic agents for immune suppression upon organ transplantation and against autoimmune disease can be developed by inhibiting the binding of CD40 with its ligand. Sawada-Hase et al., have reported that the proportion of cells strongly expressing CD40 was increased in the peripheral blood monocytes of Crohn's disease patients. However, antibodies that inhibit the binding of CD40 with its ligand have not been well understood For example, such antibodies that inhibit the binding may be effective for the functional analysis of CD40, and therapy against disease, for which activation of CD40 is required. Moreover, antibodies that inhibit CD40 ligands have been also shown to have the potential of being effective as agents against diseases with which the binding of CD40 with CD40 ligands is involved. However, it has been reported that CD40L is expressed in activated blood platelets (V. Henn et. al., Nature 391: 591, 1998). Thus, it has been reported that there is a risk of causing thrombi, if anti-CD40L antibodies are used as a therapeutic agent (T. Kawai et. al., Nat. Medi. 6: 114, 2000). From such a point of view, antibodies against CD40 can be expected to be safer than anti-CD40L antibodies, as an antibody therapeutic agent that inhibits the binding of CD40 with its ligand. Anti-CD40 antibodies are required to suppress the binding of CD40L to CD40, and not to activate CD40 by the antibody itself.
Although a huge number of studies have been conducted in the past concerning antibodies that bind specifically to human CD40 and suppress the binding of CD40L to CD40 without activating CD40, only a single case, that is a mouse anti-human CD40 antibody, named 5D12, has been reported (J. Kwekkeboom et al., Immunology 79: 439, 1993). In addition, it has not been known whether or not antibodies showing neutralization activity for B cells can also show the same for DC that is, if the antibodies can neutralize the action of CD40 ligands. Furthermore, it has been reported that the action of biotinylated anti-mouse CD40 antibodies is enhanced by cross-linking with avidin (Johnson et al., Eur J Immunol, 24: 1835, 1994). We enhanced the action of soluble CD40 ligands against a B cell line (Ramos cells) using antibodies (M2) against tags (FLAG), which had been previously provided by genetic engineering techniques to the soluble ligands, and measured the neutralization activity. Thus, we confirmed that 5D12 (ATCC No. HB-11339) exhibits only slight neutralization activity.
We have newly found that 5D12, an antagonistic antibody, has agonistic activity on its own, as a result of cross-linking even in the absence of CD40L. Conventionally, it has been reported that the action of mouse CD40 antibodies is enhanced by cross-linking of biotin with avidin (Johnson et al., Eur J Immunol, 24: 1835, 1994). Furthermore, it has been known that solid-phasing of CD40 antibodies using anti-immunoglobulin antibodies solid-phased on a plate leads to an increase in activity to suppress the proliferation of tumor cells. This has been thought to be an effect resulting from solid-phasing. However, it has not been known that when anti-immunoglobulin antibodies are added to a culture solution for cross-linking of anti-CD40 antibodies, it may become possible even for antagonistic antibodies to show agonistic activity. If antibodies to be used for therapy have antigenicity, a completely opposite effect may occur, such that antibodies which bind to CD40 antibodies in a human body are produced, and with which CD40 antibodies are cross-linked, so that activity seemingly the same as that of CD40 ligands is produced. Accordingly, in view of the safety of a therapeutic agent, it is very important to keep the antigenicity of antibodies at a low level. Consider a case wherein a therapeutic agent is developed by humanization technology based on the sequence of a variable region of a mouse antibody. Since humanized antibodies are known to have immunogenicity, anti-humanized anti-CD40 antibodies may be produced after administration. Specifically, there may be a risk that the antibodies would become agonistic antibodies. Even if the antigenicity is low, anti-CD40 antibodies may be cross-linked with antibody receptors (FcR). From these points, a preferred antagonistic antibody is a human antibody, which binds specifically to CD40, suppresses the binding of CD40L, and does not activate CD40 even by cross-linking, and exhibits weak binding to FcR.